In early June the open source search business arguably came of age when Elasticsearch closed on a $70M tranche of Series C investment funding. Elasticsearch is supported by both an open source search community and a commercial search business. In November 2012 the company received $10M in a Series A investment funding which was followed by a further $24M in a Series B funding in February 2013 -- making the total of investment funding now $104M.
Series C funding is usually seen as a substantial level of endorsement of the future of the company by the venture capital investment community. The next level, Series D, is seen as a prelude either to an Initial Public Offering or a trade sale.
To put this in perspective: Coveo has pulled in around $36M over the last few years and investments in Attivio total $42M. Intel's $740M investment in Cloudera in late March is not quite what it looks like on the surface. Also in March, BlackRock lead a $100M investment round in Hortonworks to support the development of Hadoop-based big data solutions, which might shape the future direction of the LucidWorks/Hortonworks joint venture.
Search -- in various forms -- is clearly a higher priority for the venture capital community than it seems to be for many organizations!
How to Spend $104M
The progress that Elasticsearch has already made over the last couple of years is summed up on its website as:
- Increased adoption of its three core products -- Elasticsearch, Logstash and Kibana (known as the ELK stack) -- by three times, climbing to more than 8 million total downloads
- Grown its customer base by seven times, adding Comcast, eBay, Facebook, Mayo Clinic, TomTom and Tinder to its customer roster
- Doubled its number of employees, including executives from Box, Citrix, Splunk and VMware
The company currently has just over 100 employees in nine countries. Steve Schurmann, Elasticsearch CEO, recently said in an interview with Fast Company that the company wanted the best people, regardless of location.
So where is the money going? Shay Banon, the creator of Elasticsearch, comments on the Elasticsearch site:
So, what do we plan to do with our new found venture capital? More of the same, and even more. We’ll continue to make the ELK stack open source and ever improving. Always. We’ll continue adding more commercial offerings to our product line. We’ll continue hiring great employees. And we’ll keep hiring them worldwide.”
Implications for the Search Business
The search business has changed a great deal over the last few years. First came the series of acquisitions of Exalead, Vivisimo, Endeca and Isys in 2010-2012. Now we have this series of large scale investments in companies seeking to exploit open source search code on a commercial basis. Of course it will still be possible to download Lucene, Solr, Elasticsearch and Hadoop and develop applications internally, but this is not as easy as it sounds. It takes more than Java skills to get the best from these applications, including knowing how to build and maintain the underlying server architectures and which third-party applications might be needed to enhance the core open source code.
The commercial model for open source revenue generation until recently was to offer management contracts, an approach largely pioneered by LucidWorks in its Lucid Imagination days. However support contracts do not easily scale and do not generate the margins needed to give an acceptable return on $50M+ investments. These companies need to move into the development of commercial software products and services based on open source search code that compete directly with offerings from companies such as BA-Insight, Coveo, Recommind and Sinequa, to name just a few.
Most search vendors are small companies and the extent of global support is fairly limited. It was notable from the start that the company set up offices in both Europe and the US. From Shay Banon’s statement and that of Steve Schurmann, geographic proximity to its customers, combined with a virtual global development team could act as one of the differentiators for Elasticsearch. This is important because of the comparatively low level of understanding of search implementation by the IT departments of even major corporations, and due to the complexities of dealing with multiple-language search.
It will be interesting to watch the direction in which the core code base is developed. The people who sign off on changes are the committers, many of whom are already working for companies like LucidWorks, Elasticsearch and IBM. Their employers will obviously support the committers to invest their time in ways that suit their long term objectives. There will still be independent committers, but one way or another the open source search business will now be focused on delivering a return on the investments that have been made, to the benefit of all those organizations world-wide that still cannot find the information they need to make good decisions.